It’s certainly possible that quality could be improved in various ways, but that’s an argument for increase certain quality standards and not lowering them.
Overloading existing terms with new meanings produces motte-and-bailey problems, especially when your term has slightly different meaning like the philosophers uses “principle of charity” in a different way then the people who coined the term over from philosophy of language where it has a different meaning. Nobody tries to separate the concept and thus you have a recipe for motte-and-bailey problems.
Using new words helps people to not just use old concept that are similar but make an attempt to understand the new concept that’s slightly different. Every new scientific paradigm has their own vocabulary.
You might argue that we haven’t yet a good scientific paradigm of rationality at LessWrong but I think we should aspire to it and that includes creating concepts and naming them.
It’s certainly possible that quality could be improved in various ways, but that’s an argument for increase certain quality standards and not lowering them.
Who said anything about lowering quality standards? I’m explicitly calling for increasing them, and for increasing accessibility standards as well.
Merely being obscure indicates nothing about quality.
Every new scientific paradigm has their own vocabulary
Yes. If you have genuinely new ideas, you are entitled to coin new vocabulary. But much lesswrongian vocabulary can be mapped onto mainstream vocabulary. Its an unnecessary barrier.
Who said anything about lowering quality standards?
The OP towards which I responded.
But much lesswrongian vocabulary can be mapped onto mainstream vocabulary.
Thomas Kuhn gives an example that chemists and physicists have slightly different idea of what molecule means. It’s a very similar concept but leads to different conclusions about whether an atom of helium is a molecule or not. If you talk cross disciplines this sets up misunderstands where nobody knows why the misunderstanding is there because everyone assumes they mean the same thing when they say molecule.
You avoid that by having new distinct terms. It makes it harder to engage with a discourse but it helps prevent misunderstandings.
Idea Inoculation is also worth worrying about. It was for a long time one of the reasons why CFAR didn’t release their handbook. You reduce the effects of it by using new words.
It’s also worth noting that many academic fields invent their own vocabulary even when existing terminology exists.
When it comes to the CFAR concept “tap” there’s an NLP concept called “anchor” that’s very similar and a psychology concept called “implementation intention” that’s similar as well. In the timeline the NLP concept was first, then the psychology concept came into being, then the CFAR concept came into being. Academics are usually really bad at reusing existing terms from other knowledge communities.
If someone teaches me a fourth term for the same thing I don’t think it’s a problem but it’s worth pointing out that the CFAR/NLP terms are both more practical to use as they are shorter and that matters for application. Each term also focuses on another aspect of the same concept which is interesting for it’s usage.
Psychologists also seem to invented their own word for applied rationality after us and didn’t use our term.
Most philosophical terminology has a bunch of different definitions from different authors and the terms have a lot of attached connotations.
There’s principle of charity for example which comes out of philosophy of language and often gets used outside of that with a different meaning. By using a new term like steelmanning you can move past the conflation of concepts that the philosophers engage in.
Take rationalism as you spoke about it. It’s not a term with a single meaning. There the textbook definition from Baron that’s close to the way we use the term in LessWrong. There’s an older notion of it being about abstract reasoning. Reusing that term produces a lot of trouble.
I remember someone counting the number of distinct uses of is_a and coming out at 37.
Do you have specific examples in mind where you think there’s an existing word with a single meaning?
It’s certainly possible that quality could be improved in various ways, but that’s an argument for increase certain quality standards and not lowering them.
Overloading existing terms with new meanings produces motte-and-bailey problems, especially when your term has slightly different meaning like the philosophers uses “principle of charity” in a different way then the people who coined the term over from philosophy of language where it has a different meaning. Nobody tries to separate the concept and thus you have a recipe for motte-and-bailey problems.
Using new words helps people to not just use old concept that are similar but make an attempt to understand the new concept that’s slightly different. Every new scientific paradigm has their own vocabulary.
You might argue that we haven’t yet a good scientific paradigm of rationality at LessWrong but I think we should aspire to it and that includes creating concepts and naming them.
Who said anything about lowering quality standards? I’m explicitly calling for increasing them, and for increasing accessibility standards as well.
Merely being obscure indicates nothing about quality.
Yes. If you have genuinely new ideas, you are entitled to coin new vocabulary. But much lesswrongian vocabulary can be mapped onto mainstream vocabulary. Its an unnecessary barrier.
The OP towards which I responded.
Thomas Kuhn gives an example that chemists and physicists have slightly different idea of what molecule means. It’s a very similar concept but leads to different conclusions about whether an atom of helium is a molecule or not. If you talk cross disciplines this sets up misunderstands where nobody knows why the misunderstanding is there because everyone assumes they mean the same thing when they say molecule.
You avoid that by having new distinct terms. It makes it harder to engage with a discourse but it helps prevent misunderstandings.
Idea Inoculation is also worth worrying about. It was for a long time one of the reasons why CFAR didn’t release their handbook. You reduce the effects of it by using new words.
It’s also worth noting that many academic fields invent their own vocabulary even when existing terminology exists.
When it comes to the CFAR concept “tap” there’s an NLP concept called “anchor” that’s very similar and a psychology concept called “implementation intention” that’s similar as well. In the timeline the NLP concept was first, then the psychology concept came into being, then the CFAR concept came into being. Academics are usually really bad at reusing existing terms from other knowledge communities.
If someone teaches me a fourth term for the same thing I don’t think it’s a problem but it’s worth pointing out that the CFAR/NLP terms are both more practical to use as they are shorter and that matters for application. Each term also focuses on another aspect of the same concept which is interesting for it’s usage.
Psychologists also seem to invented their own word for applied rationality after us and didn’t use our term.
Not immediately rounding down to existing terms can be very useful in research conversation as argued in https://www.lesswrong.com/posts/kvLPC5YWgSujcHSkY/how-to-play-a-support-role-in-research-conversations
Of course that doesn’t mean that every usage of a new term is useful, but the tradeoffs are complex.
I was referring more to Yudowsky’s reinvention of philosophical terminology.
Most philosophical terminology has a bunch of different definitions from different authors and the terms have a lot of attached connotations.
There’s principle of charity for example which comes out of philosophy of language and often gets used outside of that with a different meaning. By using a new term like steelmanning you can move past the conflation of concepts that the philosophers engage in.
The analytical tradition doesn’t, or at least much less so. (And noting that a situation is bad is no excuse for making it worse).
And analytical philosophy is a major rival to rationalism inasmuch as it’s a way of doing philosophy that’s based on science and logic.
But it isnt a rival to rationalism inasmuch as rationalism is a thing where one guru-like amateur philosopher solves everything.
Take rationalism as you spoke about it. It’s not a term with a single meaning. There the textbook definition from Baron that’s close to the way we use the term in LessWrong. There’s an older notion of it being about abstract reasoning. Reusing that term produces a lot of trouble.
I remember someone counting the number of distinct uses of is_a and coming out at 37.
Do you have specific examples in mind where you think there’s an existing word with a single meaning?